Print-ready version Back to the article
HMMNM Security
hmmnm.com

Governance & Compliance

ISO 27001 From Scratch Part 3: The Statement of Applicability Without Tears

By Prabhu Kalyan Samal | Published 22 November 2026 | 7 min read

https://hmmnm.com/iso-27001-from-scratch-part-3-the-statement-of-applicability-without-tears/

8 min read · 1,455 words

TL;DR: The Statement of Applicability (SoA) is the document that ties your risk register to Annex A — it decides which of the 93 controls apply to you, why, and whether they’re implemented yet. Get the justifications and status vocabulary right and your certification audit becomes a formality; get them wrong and it becomes a finding factory.

Authorization note: everything in this article is documentation work you do on your own lab machines and files. No scanning, no probing, no third-party systems. Build your ISMS in a folder, break it, and fix it — safely.

Where the SoA Sits in the ISO 27001 Machine

By Part 2 of this series you had a risk register full of risks and a mounting sense that Annex A was going to be painful. Here’s the good news: the SoA is just a table that answers four questions for each of the 93 Annex A controls.

  1. Is this control applicable to us?
  2. Why or why not? (Justification)
  3. Is it implemented yet? (Status)
  4. Where’s the proof? (Evidence reference)

Clause 6.1.3 d) of ISO 27001:2022 requires you to produce this document and compare it against Clause 6.1.3 e), which demands justification for any control you exclude. Auditors read the SoA first because it’s their map of your entire ISMS. Everything they check afterward traces back to a row in this table.

Annex A 2022: 93 Controls, 4 Themes

The 2022 revision collapsed the old 114 controls (14 domains) into 93 controls grouped into four themes. Eleven new controls were added, and many old controls were merged. Know the shape of the table:

Theme Controls Covers
A.5 Organizational 37 Policies, roles, supplier relationships, incident management, cloud services (new A.5.23)
A.6 People 8 Screening, awareness, disciplinary process, remote working
A.7 Physical 14 Perimeters, entry controls, equipment, clear desk, media handling
A.8 Technological 34 Endpoint, crypto, logging, networks, secure dev, new controls like A.8.16 (monitoring) and A.8.28 (secure coding)

Mapping Each Theme to Your Risk Register

The SoA isn’t written in a vacuum. Every “applicable” decision must trace to a risk in your register. The workflow per theme:

Keep this traceability in the register itself. Add an annex_a_refs column so the link is bidirectional: risk → control, and control → risk. In a CSV register:

id,title,annex_a_refs
R-003,No centralized logging,A.8.15
R-014,No supplier security review,A.5.19;A.5.21
R-021,Lab hardware theft,A.7.6;A.7.9

A quick consistency check from the command line — find Annex A references in the SoA that don’t appear in the risk register:

$ grep -oE 'A.[5-8].[0-9]+' soa.csv | sort -u > soa_refs.txt
$ grep -oE 'A.[5-8].[0-9]+' risks.csv | sort -u > risk_refs.txt
$ comm -23 soa_refs.txt risk_refs.txt
A.8.28

That output — a control cited in the SoA but tied to no risk — is exactly the kind of orphan an auditor finds. Either add a risk or fix the reference.

The SoA Table Anatomy

Use five columns minimum. Fancier spreadsheets add owners and target dates, but auditors care about these:

Column Contents What auditors check
Control ref & name A.5.1 Policies for information security Matches the 2022 numbering exactly
Applicability Applicable / Not applicable Consistent with justification
Justification Risk IDs + reasoning Traceable to the risk register
Status See vocabulary below Honest, current, matches evidence
Evidence Doc ID, ticket, repo path Exists and is dated

A worked row:

Control:      A.8.15 Logging
Applicable:   Yes
Justification: Risk R-003 (no centralized logging) would leave
              incidents undetectable; required for A.5.25 and
              A.5.28 response processes.
Status:       Partially implemented
Evidence:     POL-LOG-001 v1.2; jira SEC-42 (rsyslog to lab
              SIEM, due 2025-03-01)

Notice the justification names a risk ID. “Because we should” is not a justification. “Because R-003 says undetected incidents are a major risk” is.

Writing Exclusion Justifications That Pass

The 2022 standard makes exclusions easier than 2013 did — you no longer need to prove you can’t affect a control due to outsourcing. You only need a defensible reason rooted in your context, scope, and risk assessment. But weak exclusions are still the top source of major findings.

Rules That Keep Exclusions Alive

Worked Example: Cloud-Only Org Excluding Physical Controls

Imagine your scope: a two-person SaaS startup running entirely on AWS and Google Workspace, working from homes and cafés. Here’s how A.7 exclusions look in practice:

Control:      A.7.9 Security of assets off-premises
Applicable:   Not applicable
Justification: Org holds no on-premises facilities; all
              production assets run in AWS (shared-responsibility
              model, AWS handles physical security per SOC 2
              report AWS-2024-11). Portable hardware (2 laptops)
              is covered by A.7.6/A.7.10, which remain applicable
              (see rows for A.7.6, A.7.10). Residual physical risk
              tracked as R-021 and treated via full-disk
              encryption (A.8.24) and remote wipe (MDM policy
              MDM-001).
Status:       N/A — see A.7.6, A.7.10
Evidence:     Scope statement SCOPE-001 §2; AWS SOC 2 bridge
              letter on file

Why this passes: it cites the scope, names the responsible party (AWS), references the shared-responsibility evidence, and explicitly shows the risk isn’t orphaned — laptops remain covered by other applicable controls. Compare that with the version that fails:

# FAILS AUDIT:
A.7.x — Not applicable. "We are a cloud company."

Cloud companies still have laptops, phones, and home offices. Blanket theme-level exclusions are a red flag. Be surgical: exclude A.7.1–A.7.5 (facility perimeters, physical entry) if you have no facility, but keep A.7.6 (working in secure areas — yes, remote working counts), A.7.7 (clear desk — still applies at home), A.7.9, and A.7.10 (media handling) almost always.

The Status Vocabulary Auditors Check

Pick a small, defined vocabulary and never deviate. Auditors cross-check status against evidence, and creative words like “mostly done” invite scrutiny.

Status Meaning Expected evidence
Implemented Fully operating in the ISMS Live policy, config, or record, dated before audit
Partially implemented Design done, rollout incomplete Policy + rollout plan with dates
Planned Approved, not started Ticket/backlog item, target date
Not applicable Excluded with justification The justification itself + scope doc

Certification-stage reality check: for your Stage 2 audit, most controls should be Implemented. A sea of “Planned” tells the auditor your ISMS exists on paper only, which is the textbook definition of a major nonconformity. Two honest strategies exist:

Keep the SoA version-controlled so you can show its history. In your lab repo:

$ git init soa && cd soa
$ cp ../soa-master.xlsx soa-v1.xlsx 2>/dev/null; cp ../soa-master.csv .
$ git add . && git commit -m "SoA v1: 84 applicable, 9 N/A"
[master a1b2c3d] SoA v1: 84 applicable, 9 N/A
$ git tag audit-ready

Auditors occasionally ask “has the SoA changed since we last spoke?” — and git log is a beautiful answer.

Lab Exercise: Build and Validate a 93-Row SoA

Download the Annex A control list (any reputable summary of ISO/IEC 27002:2022 works), put it in a CSV, and scaffold your table:

$ awk -F, 'NR>1 {printf "%s,%s,,,%sn", $1, $2, "Planned"}' annex_a_controls.csv > soa.csv
$ wc -l soa.csv
93 soa.csv

Then iterate: for each row, search your risk register for a mapping. Rows with no risk mapping are either wrongly applicable or reveal a risk-assessment gap — both are useful findings. Finish with a completeness check:

$ cut -d, -f1 soa.csv | grep -cE '^A.[5-8].[0-9]+$'
93
$ cut -d, -f2 soa.csv | grep -cvE '^(Applicable|Not applicable)$'
0

Zero blanks means every row has an explicit decision. That alone puts you ahead of most first-time implementations.

Checklist

HMMNM Security · Independent security research and education · hmmnm.com
Printed from https://hmmnm.com/iso-27001-from-scratch-part-3-the-statement-of-applicability-without-tears/
© 2026 HMMNM Security. For personal and internal team use. Please link to the original article when sharing.